1. Field of the Invention
The present invention relates to a solid-state imaging device including light converging means above a photosensitive element, and a method of producing the same.
2. Description of the Related Art
As a type of solid-state imaging device, a charge-coupled device (CCD) type solid-state imaging device of interline transfer system is known. Such a CCD type solid-state imaging device includes a photosensitive element for converting light into electric charge and a CCD which transfers the electric charge. Referring to FIG. 4, such a CCD type solid-state imaging device 101 is illustrated. The solid-state imaging device 101 includes a plurality of photosensitive elements 1 of pn-junction photodiodes for converting light into electric charge. As is shown in FIG. 4, the plurality of photosensitive elements 1 are arranged on a semiconductor substrate 8 such as a silicon substrate, in an array of rows and columns along a first direction (hereinafter, referred to as a vertical direction) and a second direction (hereinafter, referred to as a horizontal direction) which is substantially perpendicular to the first direction. Between the photosensitive elements 1 which are adjacent to each other in the horizontal direction, vertical CCD register portions 2 extending in the vertical direction are formed. One end of each of the vertical CCD register portions 2 is connected to a horizontal CCD register portion 3 which is positioned perpendicular to the vertical CCD register portions 2. Moreover, in the vicinity of the horizontal CCD register portion 3, an output portion 4 for extracting the electric charge from the horizontal CCD register 3 is provided. Over the above-mentioned vertical CCD register portions 2 and horizontal CCD register portion 3, a light blocking film (not shown) for preventing light from reaching an area other the photosensitive elements 1.
FIG. 5A shows a cross section along the horizontal direction of the solid-state imaging device 101 taken along a line A-A' in FIG. 4. The semiconductor substrate 8 includes a photosensitive region 6 and a transfer region 5 in which the photosensitive element 1 and a CCD 31 are formed, respectively. Above the CCD 31, first and second polycrystalline silicon (polysilicon) electrodes 9 and 10 are formed. The vertical CCD register portion 2 is constituted by the CCD 31, and the first and second polysilicon electrodes 9 and 10. In addition, above the second polysilicon electrode 10, a light blocking film 11 is formed. A distance from the surface of the semiconductor substrate 8 to the top face of the light blocking film 11 is indicated by a height 15. A passivation film 12 is formed so as to cover the entire face of the substrate.
FIG. 5B shows a cross section along the vertical direction of the solid-state imaging device 101 taken along a line B-B' in FIG. 4. In the photosensitive region 6 of the semiconductor substrate 8, the photosensitive element 1 is formed. A pixel isolating region 7 is formed in the semiconductor substrate 8 between the photosensitive elements 1 which are adjacent to each other in the vertical direction. Above the pixel isolating region 7, the first and the second polysilicon electrodes 9 and 10 are formed. Above the second polysilicon electrode 10, the light blocking film 11 is formed. Then, a passivation film 12 is formed so as to cover the entire face of the substrate.
The light blocking film 11 is formed in such a manner that it covers an area which is considerably close to the photosensitive element 1 in order to reduce smear. Over the entire surface of the semiconductor substrate 8 on which the light blocking film 11 has been formed, the passivation film 12 is formed as the uppermost layer. Between the first polysilicon electrode 9 and the second polysilicon electrode 10, and between the second polysilicon electrode 10 and the light blocking film 11, an insulating film 12' such as a silicon oxide film is formed so as to insulate them from each other, respectively.
In the solid-state imaging device 101 having the above-mentioned construction, the surface of the semiconductor substrate 8 includes the transfer regions 5 and the pixel isolating regions 7 in addition to the photosensitive regions 6 having the photosensitive elements 1. According to the construction, an area occupied by the photosensitive elements 1 is small as compared with the total surface area of the semiconductor substrate 8, and among light beams 32 and 33 incident on the solid-state imaging device 101, only the light beams 32 which are incident on areas above the photosensitive regions 6 can penetrate the solid-state imaging device 101 and reach the photosensitive regions 6, so as to be converted into electric charge for providing image information. Specifically, in the case of the solid-state imaging device 101 FIGS. 5A and 5B, only approximately 25% of the incident light can contribute to provide image information. This means that the penetrating efficiency of light is not so good, and hence the sensitivity of the solid-state imaging device 101 is poor.
In order to eliminate the above disadvantage, there is proposed an additional formation of microlenses 14 on the solid-state imaging device 101 having the above-mentioned construction. The microlenses 14 can improve the penetrating efficiency of light which can reach the photosensitive regions 6, so as to enhance the sensitivity of the solid-state imaging device 101 with the microlenses 14. FIG. 6A shows a cross section along the horizontal direction of a solid-state imaging device 102 with the microlenses 14. FIG. 6B shows a cross section along the vertical direction of the solid-state imaging device 102.
In the solid-state imaging device 102 shown in FIGS. 6A and 6B, a transparent resin layer 13 is formed on the surface of the solid-state imaging device 101 as shown in FIGS. 5A and 5B, for example, and thereon, the microlenses 14 for converging the incident light onto the photosensitive regions 6 are formed. The transparent resin layer 13 is formed for the purpose of adjusting for the focal lengths of the microlenses 14. In the case where the solid-state imaging device 102 of FIGS. 6A and 6B is a color solid-state imaging device, color filters are provided above the respective photosensitive regions 6 so as to be vertically sandwiched within the transparent resin layer 13.
There are two methods for forming such a microlens. In one method, a photoresist is first applied uniformly. Part of the photoresist corresponding to the periphery of the microlens to be formed is removed by photolithography and etching techniques, so as to form a resist pattern. Then, the edge portion of the resist pattern is melted by heating, so as to form the microlens 14 ( Japanese Patent Publication No. 60-59752). As the material for the microlens 14, this method uses a material capable of being removed by the photolithography and etching techniques. The other method uses a material with no photosensitivity for the microlens 14. That is, in the other method, various kinds of materials are available as a material for the microlens 14. In the latter method, a layer of high light-transmitting material is formed and a thermally deformable resin layer is formed thereon. The thermally deformable resin layer is then selectively removed, so as to leave part of the resin layer above the photosensitive region. The remaining part of the thermally deformable resin layer is deformed by heating. By using the deformed layer as a mask, the high light-transmitting material layer is selectively removed by a dry etching technique, so as to form the microlens 14 (Japanese Laid-Open Patent Publication No. 60-53073).
When the microlens 14 is formed by either one of the above-described two methods, the heights h1 and h2 of the microlens 14 in the horizontal and the vertical cross sections of FIGS. 6A and 6B are equal to each other. However, the widths w1 and w2 are different from each other. The width w1 along the horizontal direction is obtained by subtracting a gap 16 between the microlenses 14 from the arranged pitch of the microlenses 14. The width w2 is obtained by subtracting a gap 16 from the arranged pitch of the microlenses 14 in the vertical direction. In order to efficiently converge the incident light on the photosensitive region 6, the gap 16 between the microlenses 14 should be made as small as possible, and the widths w1 and w2 of the microlens 14 should be made as large as possible. Therefore, as is shown in FIGS. 6C and 6D, a resist pattern 19 is formed on the transparent resin layer 13 in such a manner that gaps 20 between the adjacent resist patterns 19 in the horizontal and the vertical directions are set to be a minimum value which can be attained by the process. For example, in a solid-state imaging device which is designed under a submicron rule, the gap 20 is set to be 0.6 to 0.8 .mu.m. In a solid-state imaging device which is designed under a half-micron rule, the gap 20 is set to be 0.5 .mu.m.
As described above, in the solid-state imaging device 102, the heights h1 and h2 of the microlens 14 are equal, but the widths w1 and w2 thereof in the horizontal and the vertical directions are different. This means that the curvatures of the microlens 14 in the horizontal and the vertical directions are different, which causes a difference between the focal lengths in the horizontal and the vertical directions. In this specification, the term "curvature" of a microlens is defined as a multiplicative inverse of the radius of a circle which approximates the convex curve of the microlens. Accordingly, the converging conditions of the incident light 17 in the vertical and the horizontal directions are also different. Moreover, there arise the following problems except for the case where the arranged pitches of pixels are the same in the horizontal and the vertical directions. The problems are described by referring FIGS. 7A, 7B, 8, 9A, 9B, 10A, and 10B. These figures simulate the converging conditions in the solid-state imaging device with the microlens 14. The simulation utilizes Snell's law. FIGS. 7A, 9A and 10A show cross sections along the horizontal direction. FIGS. 7B, 8, 9B, and 10B show cross sections along the vertical direction. In each of the figures, the reference numeral 18 denotes a layer including the first and second polysilicon electrodes 9 and 10, the light blocking film 11, and the like above the transfer region 5 or the pixel isolating region 7. A thickness of the transparent resin layer 13 and a thickness of the microlens 14 are indicated by L and T, respectively.
FIGS. 7A and 7B show in part a solid-state imaging device for using in a NTSC (National Television System Committee) system having 525 scanning lines in the vertical direction. The solid-state imaging device has a photosensitive screen of 1/2-inch optical system. As is shown in FIG. 7A, a width of the microlens 14 in the horizontal direction is about 12.5 .mu.m, and as is shown in FIG. 7B, a width of the microlens 14 in the vertical direction is about 10.0 .mu.m. The thickness L of the transparent resin layer 13 is set to be 11 and the thickness T of the microlens 14 is set to be t1. In this case, even when the focal lengths in the horizontal and vertical directions are different due to a difference between the curvatures of the microlens 14 in the horizontal and the vertical directions, the incident light can be efficiently converged on the photosensitive region 6.
FIG. 8 shows a cross section of a solid-state imaging device for using in a
(phase alternation line) system having a photosensitive screen of 1/2-inch optical system. The solid-state imaging device has 625 scanning lines in the vertical direction. The solid-state imaging device for PAL system strongly necessitates the provision of microlenses 14, as compared with the device for NTSC system, in order not to degrade the sensitivity by reducing the pixel pitches and hence reducing an area of the photosensitive region 6. In this case, the pixel pitch in the vertical direction is set to be about 8.5 .mu.m, which is smaller than about 10.0 .mu.m in the device for NTSC system. Therefore, although the curvature of the microlens 14 in the horizontal direction is optimum, the curvature of the microlens 14 in the vertical direction is small as is shown in FIG. 8. As a result, the focal length is shorter than in the case for the NTSC system. Accordingly, part of the incident light is blocked by the light blocking film 11 in the layer 18 and cannot reach the photosensitive region 6, which causes the converging efficiency to degrade.
FIGS. 9A and 9B show cross sections of a solid-state imaging device for a NTSC system having a photosensitive screen of 1/3-inch optical system which is smaller than in the above examples. In this case, as is shown in FIG. 9B, the thickness L of the transparent resin layer 13 is reduced to be 12 (where 12&lt;11), in order to position the point of focus of the microlens 14 in the vertical direction on the photosensitive region 6.
FIGS. 10A and 10B show cross sections of a solid-state imaging device for a
system having a photosensitive screen of 1/3-inch optical system. In this case, as is shown in FIG. 10B, the thickness T of the microlens 14 is reduced to be t2 (where t2&lt;t1), in order to position the point of focus of the microlens 14 in the vertical direction on the photosensitive region 6 by increasing the focal length.
However, according to the above construction, the focal length of the microlens 14 in the horizontal direction is unavoidably increased, as is shown in FIGS. 9A and 10A. As a result, unless a height 15 (e.g., FIG. 6C) from the surface of the semiconductor substrate 8 to the top face of the light blocking film 11 is sufficiently small, the converging efficiency of the microlens 14 in the horizontal direction is degraded, because the incident light is blocked by the light blocking film 11 in the layer 18. In other words, by positioning the focus point of the microlens 14 in either one of the horizontal and the vertical directions on the photosensitive region 6, the focus point in the other direction is not positioned on the photosensitive region 6. As a result, the converging efficiency of the microlens 14 is degraded.
As described above, according to the conventional solid-state imaging device using the microlens 14, the microlens 14 may have different widths w1 and w2 in the horizontal and the vertical directions in order to maximize the light receiving area. Therefore, the focal lengths relative to the horizontal and the vertical directions of the microlens 14 are different from each other. In a solid-state imaging device with small pixel pitches for a PAL system or for a NTSC system, when the focal point in either one of the horizontal and the vertical directions is positioned on the photosensitive region, the focal point in the other direction is not positioned on the photosensitive region. As a result, there arises a problem in that the converging efficiency of the microlens is degraded.